In my experience digital zooms work in either of two ways.
#1 crop to a fraction of the pixels of the sensor (x2 = 1/1.4 width by 1/1.4 height = 1/2 of the sensor pixels, x4 =1/2 width by 1/2 height = 1/4 of the sensor pixels) and scale the pixels up to the equivalent image size of the full sensor. The processors in such cameras are "clever enough" to make the zoom factor virtually stepless. At any digital zoom factor, by calculating intermediate values the image becomes soft rather than a chequerboard. Unlike on the spy movies you cannot put detail back where there was no pixel element to detect that detail.
#2 The older cameras did not have the ability to interpolate the pixels so the file size became scaled down, the problem there is you have only certain fixed sizes so the 'zoom' was in discrete steps. Each pixel was recorded in the file without interpolation. The interpolation would happen when you went to print or display the reduced resolution image at the same size as a full sized capture. This again ends up with a soft image lacking critical detail because (except in fiction) you can't cheat the physics of too few pixels.


